3.10.1 Information score for questions of interest

Course subject(s) Module 3. Performance-based weights and the Decision Maker

A final note before we finish this module.

Recall that the information score is computed for each calibration question. For this the intrinsic range is determined, and then the information score is computed for each expert.

The same score can be computed also for each question of interest. The intrinsic range will be determined only by experts’ assessments, since the realization is obvious not available for the questions of interest.

Recall the Dutch eating habits example and the questions of interest

6) What percentage of Dutch adults will eat fruit on a daily basis by 2025?
7) How many liters of milk will be consumed on a yearly basis by the average Dutch adult in 2025?

Note that, as beforehand, the data are purely fictional.

Question Realization Expert 1
5%    50%    95%
Expert 2
5%    50%    95%
Expert 3
5%    50%    95%
6
         61      68    75       45      55       70       43     52     63
7
        67        71       80       50        61       68
     55       65       73

Recall that the overall information score of Expert 1 (for all calibration questions) is 1.62.

The information score for Question 7 is hence lower than the overall information score of Expert 1. This suggests that Expert 1 is less informative for one question of interest than for the calibration questions.

During the next module, we will investigate the consistent differences of information scores between calibration questions and questions of interest.

Creative Commons License
Decision Making Under Uncertainty: Introduction to Structured Expert Judgment by TU Delft OpenCourseWare is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Based on a work at https://online-learning.tudelft.nl/courses/decision-making-under-uncertainty-introduction-to-structured-expert-judgment//.
Back to top